Julian Hall

Just a quick query: have you considered "foo ? !bar : bar" as a way of
implementing logical xor, and if you have, why did you decide not to include
it on your page?

Slightly strange problem here; I've looked at your site and you don't seem
to have any advice on the issue, but it's thanks to your DOM tree displaying
script that I realised that IE is behaving completely inconsistently here,
so thought I'd let you know something strange is going on.  I'm using
Internet Explorer 6 on XP.

The basic problem occurs if you have a section of your DOM that was created
from HTML with overlapping tags. The particular case I'm looking at is text
copied & pasted from either [a PDF] or a broken web page into a
contentEditable 'div'.  The HTML that is copied looks something like this:

<P>Some text is here</P>
<P><STRONG>
<P>A header here</P></STRONG>
<P>Some more text</P>

The best way of seeing what happens is to drop it into a div and view it
using your DOM tree script.  It actually works pretty-much as expected. 
There's an empty <STRONG> element sitting out on the end of the second
paragraph, so if you just look at the tree you wouldn't expect "A header
here" to be displayed in bold (although it is), but otherwise it looks OK. 
The bizarre behaviour occurs if you change the script, to use the iteration
technique that I normally use.  Instead of :
           for( var x = 0; topElement.childNodes[x]; x++ ) {
               outStr += showDomTree( topElement.childNodes[x],
topCode+'_'+x );
           }

Try:
           for (var n = topElement.firstChild; n; n = n.nextSibling) {
               outStr += showDomTree( n, topCode+'_'+x )
           }

Now things appear in the wrong place and opening branches causes strange
stuff to happen.  Yet the code looks from the definition of the attributes
used like it should behave identically.

Don't know if you want to add a note to your site about this behaviour, it
could prevent somebody having the strange kinds of trouble I've been seeing
lately...